30 Jan 2023
Twitter is on fire, but is an algorithm better than a chronological timeline?
During my regular perusal of Mastodon I’ve seen a number of people publishing posts like these…
I didn’t hear about X in my timeline, that’s a failure on Mastodon’s part.
Not having an algorithm leads to information overload. It doesn’t have to be about money.
In my experience, these kinda posts tend to come from Twitter expats complaining that they prefer the algorithm that Twitter provides, and that they would like to see something similar in Mastodon.
The point around Mastodon somehow failing because a person didn’t see a particular piece of news in their timeline is, to be frank, utter bollocks.
The beauty of Mastodon’s timeline is that you can craft what you see yourself. You’re the master of your timeline’s destiny. If you missed something, that’s a failure of yours, not Mastodons.
Or you could setup lists to filter the stuff you see by topic. There really isn’t any excuse for missing stuff in Mastodon when there’s so many tools at your disposal to craft what you see on the platform.
Twitter was built from the ground up for engagement, right? So surely having an algorithm is the way to go if you want true engagement online.
The problem is, that argument is flawed. I think Twitter was built from the ground up to make money, and to do that, they need more eyes on the platform. But how do they get more eyes?
Well, in my opinion, they do it by promoting provocative posts that raise your blood pressure and get you to stick around. There’s also a fair number of people who manipulate the algorithm, just like on Google.
So you end up with a timeline that’s littered with both provocative posts, and posts from people who know how to market themselves on Twitter. Because of all this, you do not get genuine human engagement on Twitter. Not to the same level as Mastodon, at least.
There’s lots of examples online where folk with many more followers on Twitter, post the the same thing to both platforms, and get far more engagement on Mastodon.
For example, this post by Kate was published to both Mastodon and Twitter. She has comparable follower numbers on both platforms (1.6k on Masto, 1.2k on Twitter). The post garnered 12 likes on Twitter and 1,700 like on Mastodon.
That isn’t a typo. She received more than 1,000x the engagement on Mastodon, than Twitter.
Tell me about that algorithm that was designed from the ground up for engagement again…
Here’s another example from Iconfactory. They made the Twitteriffic Twitter app, so this should be a great example, considering they created one of the best 3rd party apps for Twitter.
On Twitter, they received 9 likes and 3 retweets. On Mastodon the post received 287 favourites (likes) and 163 boosts (retweets). That’s many orders of magnitude more, and they have 4x the amount of followers on Twitter.
But Kev, this is from a few days ago and the Twitter algorithm needs time to work…
– Twitter fans everywhere
Ok, it’s been a few days since those posts were published. Maybe the Twitter algorithm caught up and things are different now?
Hahaha noooo, of course they aren’t. They’re worse. As of right now those posts have 8 retweets and 16 likes on Twitter, versus 476 favourites and 266 boosts on Mastodon.
Please bear in mind here, that the post contains bad news for Mastodon folks. The Inconfactory folks are basically saying that they won’t be building a Mastodon app anytime soon, because so many good ones already exist.
With this being bad news for Mastodon, you would assume the engagement would be way down. But no, it’s way up.
Why is this? Well, dear reader, it’s because Mastodon is built by people, not algorithms.
These aren’t unique scenarios, either. Here’s another example, and another, and another, and another, and another, and another, and finally another from Brian Krebs. Are you starting to get the picture?
Just want to quickly thank all the fine people of Mastodon for sending me these links - I couldn’t remember them all.
My preference for a chronological timeline isn’t about concern for companies making money from me. It’s about concern for my time.
I can login, read some posts, then logout. If I want to pickup where I left off later, I know all the posts will be there, in the same order. They won’t have disappeared into the ether, just because an algorithm deemed them unimportant.
Having a chronological timeline also prevents doomscrolling. If I refresh the timeline, I’m not presented with a raft of new posts that have been computationally stitched together.
Oh no, dear reader. Instead, I get the posts that have been published in the interim between now and the last time I refreshed the timeline. That could be a hundreds of posts if I’ve been offline a few days, or just a handful if it’s been a few minutes.
Finally, as I touched on earlier, because everything is chronological and there’s no algorithm at work, I can manipulate my timelines as I see fit. I can add filters, follow hashtags, create lists and more. All safe in the knowledge that whatever is in there, I have control over it and some random piece of code isn’t manipulating things on my behalf. For, you know, engagement reasons…
I think it’s clear by now that I have a strong preference for a chronological timeline. But given the examples above (of which there are many more) I’d argue that having a chronological, user-curated timeline is far more useful for engagement purposes than an algorithm which seemingly arbitrarily decides on what it should feed you.
I know some people will disagree, and that’s your prerogative. Please feel free to write your own response; but remember, don’t bother posting it to Twitter, as it will likely disappear into the big blue ether.
Alternatively, you can email me using the button below. There’s no algorithm at work in my inbox. Pinky promise. 🙃